Language Tools and Resources for Polish

This page contains a list of publicly available language tools and resources.

* NKJP-PodkorpusMilionowy-1.0.tgz, the manually annotated 1-million worsd subcorpus of the NJKP, available on GNU GPL v.3, * NKJP-PodkorpusMilionowy-1.1.tgz, the manually annotated 1-million word subcorpus of the NJKP, available on GNU GPL v.3, * NKJP-PodkorpusMilionowy-1.0-poliqarp-bin.tgz, the binary version of corpus to be used with standalone Poliqarp tool.

Written corpora of contemporary Polish

• National Corpus of Polish (NKJP),

• PWN Corpus,

• National Photocorpus of Polish (NFJP),

• Dictionaries as Corpora,

• Polish Parliamentary Corpus,

• Polish Summaries Corpus,

• PICLE corpus, the Polish sub-corpus of the International Corpus of Learner English (ICLE),

• IMPACT ground-truth data for selected Polish historical documents from PIONIER Digital Libraries Federation,

  • Now available also as corpora in the Poliqarp for DjVu search engine,

• KPWr, Polish Corpus of Wrocław University of Technology, collection of documents available on Creative Common license annotated with syntactic chunks, proper names, semantic relations, anaphora and word senses,

• Polish Corpus of Suicide Notes,

• Polish Wikipedia Corpus,

• gpwEcono, a corpus of stock market reports, with manual word sense annotation,

• plWikiEcono, a corpus of Polish Wikipedia articles from the domain of economy,

• Polish Coreference Corpus, a corpus of Polish coreference relations, created as part of the CORE project,

• ArgDB-pl, a Polish corpus of arguments in natural contexts,

• PIEWiC, a Polish corpus of errors automatically extracted from Wikipedia revisions,

• PolEval, corpora and other text resources created for PolEval shared tasks,

• Polish PARSEME corpus, annotated manually for verbal multiword expressions in 20 languages including Polish, used in the PARSEME shared task 1.1; the Polish subcorpus is aligned with automatic dependency annotations in the UD format (A. Savary),

• MweLitRead - a corpus of literal readings of Polish verbal MWEs stemming from the Polish PARSEME corpus (A. Savary, S. Cordeiro),

• PELCRA Learner English Corpus (PLEC),

• Polish text corpora included in Sketch Engine,

• Microcorpus of Synesthetic Metaphors.

Written corpora of historical Polish

• eFontes Mediae et Infimae Latinitatis Polonorum (1000–1550, IJP PAN)

• Corpus of old Polish (up to 1500) (IJP PAN)

• 15. century New Testament translations (IJP PAN)

• IMPACT project corpus (1570–1756, KLF UW)

• Corpus of 16. century Polish (IBL PAN)

• PolDiLemma, the Middle Polish Diachrone Lemmatised Corpus (16–18th c., R. Meyer)

• PolDi, a Polish Diachronic Online Corpus (R. Meyer)

• KORBA, electronic corpus of 17th and 18th century Polish texts (1601–1772, IJP PAN)

• Corpus of the 19. century Polish (1830–1918, IJP UW)

• Manually annotated and transcribed corpus of the 19th century Polish, (1830–1918, IPI PAN)

• ChronoPress, corpus of press texts from 1945–1954 (A. Pawłowski),

• Polish language of the 1960s / Frequency corpus (I. Kurcz, A. Lewicki, J. Sambor, J. Woronczak, K. Szafran, J. S. Bień, M. Woliński).

Corpus-related tools and resources

• Poliqarp, a corpus indexing and search engine (please see also the beta version of Poliqarp 1.1 and 1.3 with statistical extensions and several corpora indexed with Poliqarp 2),

• Anotatornia, a system for multi-level manual annotation of corpora,

• Anotatornia2, new version of Anotatornia geared towards annotation of historical corpora,

• Inforex, a web-based system designed for managing and annotating text corpora on the semantic level,

• Smyrna, a simple, light-weight Polish concordancer,

• korpusy.net, a corpus research-related website (B. Gałkowski),

• Korpusomat, a tool for creation of searchable own corpora (Ł. Kobyliński, W. Kieraś).

Spoken corpora

• The PELCRA conversational corpus of Polish: approx. 2.2 million words of casual conversational spoken Polish collected and processed in the years 2001-2015 in a number of research projects, including PELCRA, NKJP, CESAR and CLARIN-PL. Available under CC-BY-NC. All transcriptions can be accessed through the Spokes web interface and programmatically through a REST API.

• The annotated corpus of spoken dialogues (LUNA project, corpus data available at the end of the page),

• AGH speech corpus, around 9 hours, word-annotated Polish speech corpus (AGH DSP),

• Audiovideo corpus of Polish speech (AGH DSP),

• NKJP search engine for spoken-conversational data,

• Acoustic database for Polish unit selection speech synthesis (ELRA resources),

• Acoustic database for Polish concatenative speech synthesis (ELRA resources),

• Corpus of emotions in speech (AGH DSP).

Language models

• N-grams from the balanced subcorpus of the National Corpus of Polish,

• N-gram model of Polish (AGH DSP),

• Distributional semantic models trained on orthographical, lemmatized word forms (A. Wawer),

• Even more distributional semantic models based on NKJP (A. Mykowiecka, M. Marciniak, P. Rychlik),

• Wikipedia-based word embeddings for Polish (M. Rogalski, P. Szczepaniak),

• Wikipedia2Vec – pretrained embeddings for Polish (I. Yamada, A. Asai, H. Shindo, H. Takeda., Y Takefuji),

• Slavic BERT NER (see also Deep Pavlov website),

• RoBERTa and links many other useful resources (S. Dadas),

• Polbert (D. Kłeczek),

Parallel corpora and translation memories

• ParaSol, a parallel corpus of Slavic and other languages,

• PolUKR, a Polish-Ukrainian parallel corpus,

• OPUS, an open source parallel corpus (European Parliament, EMEA, KDE, movie subtitles),

• "1984", an annotated parallel corpus of George Orwell's "1984" in 15 languages, MULTEXT-East, v.4 (licensed download),

• InterCorp, a multilingual parallel corpus,

• Leeds collection of Internet corpora,

• LAGUN corpus,

• PELCRA Parallel corpora, a collection of downloadable parallel corpora available under the CC-BY and CC-BY-NC licensed developed by the PELCRA team

• Paralela Polish-English corpus

• JRC-Acquis Multilingual Parallel Corpus,

• PSI collection of parallel corpora, a growing collection of parallel corpora pairing Polish with other european languages,

• Polish-Russian Parallel Corpus,

• MyMemory, freely available multilingual TM,

• TAUS Data, a multilingual TM from the members of TAUS Data Association,

• Glosbe, an open source TM.

• AlphaMWE Parallel English-Chinese, English-Polish and English-German parallel corpus annotated with multiword expressions

Machine-readable dictionaries

• plWordNet, Polish WordNet, Słowosieć (M. Piasecki),

• POLNET, another Polish Wordnet (Z. Vetulani),

• Polish OpenThesaurus, słownik synonimów – a crowdsourced Polish thesaurus (M. Miłkowski),

• Słownik języka polskiego (d. alternatywny), Polish ispell dictionaries, along with some definitions and online form display,

• Nowy słownik angielsko-polski (T. Piotrowski, Z. Saloni),

• Polish OpenCYC (A. Pohl),

• Polish machine-generated dictionaries, available on Creative Commons (J. Kazojć),

• List of all Polish surnames, licence unknown, see further information on this resource,

• Gazetteer for Polish Named Entities (A. Savary, M. Lenart, J. Piskorski),

• Triggers for Polish Named Entities (M. Baron, L. Manicki, A. Savary),

• NELexicon contains more than 1.4 millions of proper names (M. Marcińczuk, A. Musiał, M. Janicki),

• Walenty, the Polish Valence Dictionary (E. Hajnicz, W. Kieraś, A. Patejuk, A. Przepiórkowski, F. Skwarski, M. Świdziński, M. Woliński),

• Syntatic-generative dictionary of Polish verbs (K. Polański),

• SAWA, the Grammatical Lexicon of Warsaw Urban Proper Names (M. Marciniak, C. Heliasz, J. Rabiega-Wiśniewska, P. Sikora, M. Woliński, A. Savary),

• SEJF, the Grammatical Lexicon of Polish Phraseology (M. Czerepowicka, A. Savary),

• SEJFEK, the Grammatical Lexicon of Polish Economical Phraseology (F. Makowiecki, A. Savary),

• WikiTopoPl, a multilingual lexicon of 155,000 Polish geographical proper names extracted from Wikipedia and their equivalents in Bulgarian, Croatian, English, German, modern Greek, Hungarian, Romanian, Serbian and Slovak (L. Manicki),

• Prolexbase 2.0, a multiligual relational dictionary of proper names in Polish, English and French (M. Baron, B. Bouchou-Markhoff, L. Manicki, D. Maurel, A. Savary, M. Tran),

• DeepER Entity Library, a database containing around 900,000 entities, each described by its textual representations in Polish (names) and WordNet synsets.

Human-readable dictionaries

• Słownik gramatyczny języka polskiego,

• Wielki Słownik Języka Polskiego,

• Słownik języka polskiego PAN pod red. W. Doroszewskiego,

• Wikisłownik,

• Słownik wyrazów obcych i zwrotów obcojęzycznych Władysława Kopalińskiego,

• Słownik synonimów i antonimów Piotra Żmigrodzkiego,

• Słownik polszczyzny XVI wieku,

• Elektroniczny słownik języka polskiego XVII i XVIII wieku,

• Poliqarp for DjVu search engine for J. Karłowicz, A. Kryński, W. Niedźwiedzki. Dictionary of Polish. Warsaw 1900–1927,

• Poliqarp for DjVu search engine for S. Bąk, M. R. Mayenowa, F. Pepłowski (eds.). Dictionary of the 16th century Polish. Wrocław — Warszawa, 1966-???? (work in progress),

• Poliqarp for DjVu search engine for M. Samuel Bogumił Linde. Dictionary of Polish (2nd edition). Lwów 1854-1861,

• Poliqarp for DjVu search engine for B. Chlebowski, F. Sulimierski, W. Walewski (eds.), The Geographical Dictionary of the Polish Kingdom and other Slavic Countries, Warszawa 1880-1902,

• Edycja elektroniczna Słownika wileńskiego,

• PELCRA HASK Collocation Dictionaries generated for Polish and English,

• Słownik ukraińsko-polski pod redakcją Janusza A. Riegera. Materiały do słownika: Litera „O”.

Morphological tools and resources

• SGJP, Grammatical Dictionary of Polish (the list of inflected forms is available with Morfeusz),

• PoliMorf, an inflectional dictionary of Polish,

• Morfeusz SGJP, morphological analyser,

• Index a tergo of Polish word forms (J. Tokarski, Z. Saloni),

• Morfologik, morphological analyser (M. Miłkowski, D. Weiss),

• SAM, morphological analyser (K. Szafran),

• UAM Text Tools (P. Obrębski, Z. Vetulani; see also http://utt.wmi.amu.edu.pl/trac/wiki/),

• MULTEXT-East, v.4, morphosyntactic specifications and documentation for 16 languages,

• Sample morphosyntactic Polish lexicon, the MULTEXT-East morphosyntactic lexicons,

• KIPI->MTE, a converter from TaKIPI to MULTEXT-East morphosyntactic format (A. Radziszewski, N. Kotsyba),

• MACA, Morphological Analysis Converter and Aggregator (A. Radziszewski, T. Śniatowski),

• Lexical analyser and a Polish proof-reader (S. Galus),

• Neurosoft Gram (demo of a morphological analyser),

• Baza fleksyjna języka polskiego, inflection database of Polish words (W. Lubaszewski, B. Moskal, P. Pietras, P. Pisarek, T. Rokicka),

• Finite state utilities (J. Daciuk),

• Stempel, another stemmer (A. Białecki),

• WCCL, toolkit for morphosyntactic feature generation (A. Radziszewski, A. Wardyński, T. Śniatowski, P. Kędzia),

• LemmaGen, Multilingual Open Source Lemmatisation for 11 EU languages, including Polish (M. Jursic, T. Erjavec et al.),

• LemmaPL, a lemmatization tool for Polish.

Taggers

• TaKIPI, a morphosyntactic tagger for Polish (Decision Trees),

• PANTERA, a morphosyntactic tagger for Polish (Transformation-Based Learning),

• WMBT, a morphosyntactic tagger for Polish (Memory-Based Learning),

• TaCo, a statistical morphosyntactic tagset converter for positional tagsets (e.g. Polish),

• WCRFT, a morphosyntactic tagger for Polish (Conditional Random Fields),

• Concraft, a morphosyntactic disambiguation tool for Polish (Constrained Conditional Random Fields),

• PoliTa, a morphosyntactic meta-tagger,

• KRNNT, a morphological tagger for Polish based on recurrent neural networks,

• NKJP model for TnT Tagger, a trained model usable on Morfeusz-segmented text with TnT Tagger,

• A PoS tagger trained on the 1M NKJP corpus and using Morfeusz (Pęzik & Laskowski 2017) with a REST API.

Parsers, grammars, treebanks

• PDB 2.0, a dependency treebank of Polish (A. Wróblewska),

• PDB-UD, a version of PDB 2.0 in Universal Dependencies format (A. Wróblewska),

• PDBparser, a Polish dependency parser (A. Wróblewska),

• Składnica, a hybrid constituency/dependency treebank of Polish,

  • Składnica main page,

  • SkładnicaMWE, a constituency version of Składnica with multiword expression annotations (J. Waszczuk, A. Savary),

  • Składnica search engine (M. Woliński),

• TAG grammar of Polish,

• POLFIE, an LFG grammar of Polish

  • POLFIE as a web service.

• Świgra, a DCG parser,

  • On-line demo,

• Spejd, a shallow parsing and disambiguation system,

  • the current version of the system,

  • Spejd grammar of Polish (version 1.0), developed by K. Głowińska within NKJP, available on GNU GPL v.3,

  • Spejd grammar of Polish with lemmatisation of Polish nominal syntactic groups,

  • SEJFEK4Spejd - a Spejd grammar version of SEJFEK and a converter from dictionary to grammar,

• Dendrarium, a treebank development system,

• A Treebank / Test Suite for Polish,

• Analizator syntaktyczny AS (M. Woliński),

• Formalny opis składniowy zdań polskich (S. Szpakowicz),

• Serwer LAS / Linguistic Analysis Server,

• Disaster (DISAmbiguator and STatistical chunkER) – a Python module for chunking and morphosyntactic disambiguation,

• Iobber, a CRF chunker for Polish,

• Krzaki (bushes), a manually annotated for dependency structure 20k-sentence corpus of Polish.

• ENIAM (W. Jaworski).

Semantic resources

• CDSCorpus, a dataset of 10k pairs of Polish sentences manually annotated for semantic relatedness and entailment (A. Wróblewska)

• Probing datasets, Polish and English probing datasets for linguistic verification of sentence embeddings (A. Wróblewska)

Sentiment analysis, opinion mining

• Polish sentiment dictionary, with sentiment scores computed using supervised methods (A. Wawer),

• Polish Linguistic Category Model following a typology of verb categorization in terms of their abstractness, also a tool to measure language abstraction (A. Wawer),

• Polish dependency treebank with sentiment annotations (A. Wawer),

• HateSpeech corpus, 2000 manually annotated documents representing various types and degrees of offensive language expressed toward minorities,

• Sincerity Corpus (Korpus Szczerości), a collection of fake and real reviews,

• you can also test Sentipejd – sentiment analysis tool in the Multiservice (please select a tagger first),

• Nencki Affective Word List.

Coreference

• Polish Coreference Corpus, a 500 M corpus of general nominal coreference in Polish (M. Ogrodniczuk),

• Polish Coreference Tools, a suite of Polish coreference resolution tools, created as part of the CORE project.

Speech analysis and synthesis tools

• Skrybot, commercial speech recognition system (L. Pawlaczyk, P. Bosky),

• Ivona, commercial text-to-speech system (Expressivo),

• Techmo TTS demo (Techmo),

• Vocalizer, commercial text-to-speech system (Nuance),

• Acapela, text to speech demo,

• Synteza mowy polskiej, automatic speech recognition and speech synthesis demos, with background information (K. Szklanny),

• System syntezy mowy ciągłej (G. Demenko, S. Grocholewski),

• Polish MBROLA database (K. Szklanny, K. Marasek),

• SynTalk, commercial speech synthesis system (NeuroSoft),

• PrimeSpeech, commercial speech recognition systems,

• OrtFon, phonetic transcriber (AGH DSP),

• Sarmata, automatic speech recognition system for Polish (AGH DSP),

• Anotator, speech corpora anotator dedicated for Polish and focused on connecting existing resources (AGH DSP),

• System rozpoznawania mówcy (AGH DSP).

Machine translation demonstrations

• iTranslate4.eu (multiple languages, allows comparing translation engines),

• Bing Translator (multilingual),

• Google Translate (multilingual),

• InterTran (multilingual),

• LingvoBit (EN-PL-EN),

• Systran (EN-PL, PL-FR and some more),

• Esperantilo (integrated Esperanto editor, with MT for EO-PL-DE-EN-SV),

• Thetos (PL-Sign language).

Summarizers

• Lakon, a system for news summarization (A. Dudczak),

• PolSum (S. Kulików),

• Summar (Ł. Pawluczuk),

• Summarizer (J. Świetlicka),

• you can also test Lakon, Open Text Summarizer and Summarizer in Multiservice

• and take a look at the Polish Summaries Corpus.

Diacritization

• Poliszynel (P. Sawicki),

• spolszcz.pl (P. Sawicki),

• Polonizator (TiP),

• Polonizer,

• fsa_accent (J. Daciuk),

• pliterki (W. Muła).

Named entity recognition

• Nerf, a tool for named entity recognition, available on GNU GPL v.3 (J. Waszczuk),

• Liner2, named entity recognizer released on GNU GPL with models to recognize 5 and 56 categories of proper names (M. Marcińczuk and M. Janicki),

• TIMEX, a model for Liner2 to recognize and normalize temporal expressions (J. Kocoń and M. Marcińczuk).

Multiword expression software

• TermoPL, multiword expression extraction tool,

• VMWE identifiers, systems having participated in the PARSEME shared task for automatic identification of verbal MWEs, 13 out of 17 systems submitted results for Polish,

• PARSEME-FR demonstrator, including the ATILF-LLF multiword expression identifier for Polish,

Aggregating services

• Multiservice, a sample interface for running NLP Web services for Polish (see also usage and format),

• Online demos of tools for processing Polish texts (CLARIN-PL),

• PSI-Toolkit, a chain of publicly available tools for automatic processing of Polish.

Other

• Mobile plWordNet, free mobile application for plWordNet browsing (J. Kocoń), * Kolokacje, a Web crawler and collocation finder (A. Buczyński),*/

• WSDDE, a system for designing and performing Word Sense Disambiguation experiments (R. Młodzki et al.),

• Frazeo, a search engine and clusterer of news in Polish (P. Pęzik),

• Segment, a rule-based sentence tokenizer supporting SRX standard (J. Lipski; the Polish rules are available in LanguageTool project, see here for short instructions on how to use the tool),

• Toki, a tokenizer supporting SRX standard, C++ library and toolkit (T. Śniatowski and A. Radziszewski),

• Translatica SRX sentence segmentation rules for Polish (LGPL),

• SyMGIZA++, an extension of Giza++ that computes symmetric word alignment models,

• Hipisek, an experimental question answering system (M. Walas),

• Narzędzia dygitalizacji tekstów, Poliqarp for DjVu i inne programy,

• Fextor, a feature extraction framework,

• LexCSD, a system for semi-automatic sense disambiguation,

• SuperMatrix, a general tool for lexical semantic knowledge acquisition,

• WordnetLoom, an wordnet editor application,

• Toposław, tool for the creation of electronic inflectional dictionaries of multi-word units,

• CorpCor, a web-based tool for correcting morphosyntactic annotation in TEI XML encoded corpora (e.g. NKJP).

• Stylo 2, stylometry demo,

• DeepEvents, event extraction in Polish, based on deep neural networks.

• Word similarity, calculation of the similarity of words based on word embeddings, on-line service,

• Baltoslav, with several script converters (Romanizer, Cyrillizer, IPA Converter etc.),

• SpacyPL, Polish language models and resources for Spacy

• Jasnopis, analyzer of text obscurity level

• AIDe, corpus of image descriptions in Polish (A. Wróblewska)